[edit]
1 一危 #
![[http]](/moniwiki/imgs/http.png)
EXCEL 譟一覦覯 覦 糾覿襯 伎.
cname <- c("ID", "蟲襷る", "磯","碁一", "碁", "覦覓碁", "蟇一朱")
x = read.table("c:\\data\\disc.txt", col.names = cname)
head(x)
disc.txt> head(x) ID 蟲襷る 磯 碁一 碁 覦覓碁 蟇一朱 1 1 A 48 9000 4 5 6 2 2 A 58 8000 6 4 20 3 3 A 52 7000 6 4 12 4 4 A 63 7000 6 4 15 5 5 A 59 8000 4 6 6 6 6 A 38 11000 5 4 10 >
蟲襷る A, B覲襦 蠏(mean)螻 覿(var) 螻 苦朱 螳螳 觚 襷 螻壱.
x1 <- subset(x, 蟲襷る=="A") x2 <- subset(x, 蟲襷る=="B")
[edit]
2 殊 覲 #
library("MASS")
rs<-lda(蟲襷る~磯+碁一+碁+覦覓碁+蟇一朱, data=x)
rs
蟆郁骸
> rs
Call:
lda(蟲襷る ~ 磯 + 碁一 + 碁 + 覦覓碁 +
蟇一朱, data = x)
Prior probabilities of groups:
A B
0.5 0.5
Group means:
磯 碁一 碁 覦覓碁 蟇一朱
A 55.0 8400 4.8 4.6 9.3
B 32.9 5400 2.3 1.9 2.5
Coefficients of linear discriminants:
LD1
磯 -0.0982250865
碁一 -0.0008926856
碁 -0.7287341389
覦覓碁 -0.6346846738
蟇一朱 -0.0578973013
>
[edit]
3 蟆郁骸伎 #
p<-predict(rs) x <- cbind(x, p$x) x
蟆郁骸
> x ID 蟲襷る 磯 碁一 碁 覦覓碁 蟇一朱 LD1 1 1 A 48 9000 4 5 6 -3.716870 2 2 A 58 8000 6 4 20 -5.439781 3 3 A 52 7000 6 4 12 -3.494566 4 4 A 63 7000 6 4 15 -4.748734 5 5 A 59 8000 4 6 6 -4.539345 6 6 A 38 11000 5 4 10 -4.845629 7 7 A 49 9000 5 5 9 -4.717521 8 8 A 62 8000 3 7 5 -4.682073 9 9 A 55 9000 5 3 5 -3.805913 10 10 A 66 8000 4 4 5 -3.899654 11 11 B 28 6000 2 2 2 4.518800 12 12 B 33 4000 3 3 3 4.391730 13 13 B 26 5000 3 1 2 5.513887 14 14 B 40 7000 2 2 1 2.505311 15 15 B 31 5000 2 1 3 5.693598 16 16 B 24 6000 2 1 3 5.488488 17 17 B 47 6000 3 1 1 2.616372 18 18 B 30 7000 1 2 3 4.100501 19 19 B 40 4000 1 2 2 5.854205 20 20 B 30 4000 4 4 5 3.207192 >
LD1螳 A觚襯 蟲襷ろ 願, B觚襯 蟲襷ろ . 覲り 覲企 . れ ろ企慨覃 螻磯 蟆郁骸螳 15.46785襯 蠍一朱 る 蟆 .
calc <- with(x, 磯 * -0.0982250865 + 碁一 * -0.0008926856 + 碁 * -0.7287341389 + 覦覓碁 * -0.6346846738 + 蟇一朱 * -0.0578973013) x <- cbind(x, calc) calc2 <- with(x, LD1 - calc) x <- cbind(x, calc2) x
蟆郁骸
> x ID 蟲襷る 磯 碁一 碁 覦覓碁 蟇一朱 LD1 calc calc2 1 1 A 48 9000 4 5 6 -3.716870 -19.184718 15.46785 2 2 A 58 8000 6 4 20 -5.439781 -20.907629 15.46785 3 3 A 52 7000 6 4 12 -3.494566 -18.962415 15.46785 4 4 A 63 7000 6 4 15 -4.748734 -20.216583 15.46785 5 5 A 59 8000 4 6 6 -4.539345 -20.007193 15.46785 6 6 A 38 11000 5 4 10 -4.845629 -20.313477 15.46785 7 7 A 49 9000 5 5 9 -4.717521 -20.185369 15.46785 8 8 A 62 8000 3 7 5 -4.682073 -20.149922 15.46785 9 9 A 55 9000 5 3 5 -3.805913 -19.273761 15.46785 10 10 A 66 8000 4 4 5 -3.899654 -19.367502 15.46785 11 11 B 28 6000 2 2 2 4.518800 -10.949048 15.46785 12 12 B 33 4000 3 3 3 4.391730 -11.076119 15.46785 13 13 B 26 5000 3 1 2 5.513887 -9.953962 15.46785 14 14 B 40 7000 2 2 1 2.505311 -12.962538 15.46785 15 15 B 31 5000 2 1 3 5.693598 -9.774251 15.46785 16 16 B 24 6000 2 1 3 5.488488 -9.979361 15.46785 17 17 B 47 6000 3 1 1 2.616372 -12.851477 15.46785 18 18 B 30 7000 1 2 3 4.100501 -11.367347 15.46785 19 19 B 40 4000 1 2 2 5.854205 -9.613644 15.46785 20 20 B 30 4000 4 4 5 3.207192 -12.260657 15.46785
碁襦 覲 .
cname <- c("ID", "蟲襷る", "磯","碁一", "碁", "覦覓碁", "蟇一朱")
x = read.table("c:\\data\\disc.txt", col.names = cname)
rs<-lda(碁~磯+碁一+蟲襷る+覦覓碁+蟇一朱, data=x)
rs
p<-predict(rs)
x <- cbind(x, p$x)
x
蟆郁骸
> rs
Call:
lda(碁 ~ 磯 + 碁一 + 蟲襷る + 覦覓碁 +
蟇一朱, data = x)
Prior probabilities of groups:
1 2 3 4 5 6
0.10 0.20 0.20 0.20 0.15 0.15
Group means:
磯 碁一 蟲襷るB 覦覓碁 蟇一朱
1 35.00000 5500.000 1.00 2.00 2.50000
2 30.75000 6000.000 1.00 1.50 2.25000
3 42.00000 5750.000 0.75 3.00 2.75000
4 50.75000 7250.000 0.25 4.75 5.50000
5 47.33333 9666.667 0.00 4.00 8.00000
6 57.66667 7333.333 0.00 4.00 15.66667
Coefficients of linear discriminants:
LD1 LD2 LD3 LD4
磯 0.0108341882 -0.0842979470 -0.0182453017 0.0079837572
碁一 -0.0005159674 0.0003233735 0.0005008073 0.0005652328
蟲襷るB -3.5186494778 -3.6997526685 1.3377932310 5.0233382564
覦覓碁 -0.5541865385 -0.0619435929 -0.5560926546 0.7505362060
蟇一朱 0.5078180145 -0.1710574267 0.1361173174 0.1627906244
LD5
磯 -0.1117330322
碁一 -0.0005554468
蟲襷るB -3.7199816354
覦覓碁 0.0469918970
蟇一朱 -0.0118410355
Proportion of trace:
LD1 LD2 LD3 LD4 LD5
0.8430 0.0983 0.0578 0.0006 0.0004
> x
ID 蟲襷る 磯 碁一 碁 覦覓碁 蟇一朱 LD1 LD2 LD3 LD4 LD5
1 1 A 48 9000 4 5 6 -0.1993729 2.06204706 -0.6506451 0.03737132 0.3220854
2 2 A 58 8000 6 4 20 8.0885751 -1.43716633 1.1278296 1.08050866 -0.4525645
3 3 A 52 7000 6 4 12 4.4769932 0.11370722 -0.3524444 -0.83495164 0.8680088
4 4 A 63 7000 6 4 15 6.1196233 -1.32674248 -0.1447908 -0.25875844 -0.3965776
5 5 A 59 8000 4 6 6 -0.1184160 0.74945251 -1.9082434 0.31049609 -0.3045392
6 6 A 38 11000 5 4 10 1.2458091 2.92948751 1.6339845 0.98862558 0.2341661
7 7 A 49 9000 5 5 9 1.3349153 1.46457684 -0.2605384 0.53372695 0.1748293
8 8 A 62 8000 3 7 5 -1.1479180 0.60567250 -2.6551893 0.92219294 -0.5809054
9 9 A 55 9000 5 3 5 0.4770215 1.76690605 0.1977058 -1.57060541 -0.5421886
10 10 A 66 8000 4 4 5 0.5579784 0.45431149 -1.0598925 -1.29748065 -1.1688132
11 11 B 28 6000 2 2 2 -2.7555165 -0.05180681 0.6734409 0.30256502 0.4094934
12 12 B 33 4000 3 3 3 -1.7157794 -1.35304466 -0.8393757 0.12534510 0.9968727
13 13 B 26 5000 3 1 2 -1.7070310 -0.14464087 0.7652168 -1.02917147 1.1414143
14 14 B 40 7000 2 2 1 -3.6492916 -0.56895121 0.8191873 0.80081225 -1.4749088
15 15 B 31 5000 2 1 3 -1.1450420 -0.73718803 0.8101076 -0.82646206 0.5709081
16 16 B 24 6000 2 1 3 -1.7368487 0.17627114 1.4386320 -0.31711559 0.7975926
17 17 B 47 6000 3 1 1 -2.5032984 -1.42046679 0.7467555 -0.45907043 -1.7485851
18 18 B 30 7000 1 2 3 -2.7419975 -0.06808659 1.2738749 1.04655592 -0.3812606
19 19 B 40 4000 1 2 2 -1.5935715 -1.71012927 -0.5471174 -0.73209543 0.1795906
20 20 B 30 4000 4 4 5 -1.2868324 -1.50420926 -1.0684978 1.17751128 1.3553816
[edit]
4 iris #
tmp <- iris
library("MASS")
rs<-lda(Species~Sepal.Length+Sepal.Width+Petal.Length+Petal.Width, data=tmp)
rs
蟆郁骸
> rs
Call:
lda(Species ~ Sepal.Length + Sepal.Width + Petal.Length + Petal.Width,
data = tmp)
Prior probabilities of groups:
setosa versicolor virginica
0.3333333 0.3333333 0.3333333
Group means:
Sepal.Length Sepal.Width Petal.Length Petal.Width
setosa 5.006 3.428 1.462 0.246
versicolor 5.936 2.770 4.260 1.326
virginica 6.588 2.974 5.552 2.026
Coefficients of linear discriminants:
LD1 LD2
Sepal.Length 0.8293776 0.02410215
Sepal.Width 1.5344731 2.16452123
Petal.Length -2.2012117 -0.93192121
Petal.Width -2.8104603 2.83918785
Proportion of trace:
LD1 LD2
0.9912 0.0088
>
Proportion of trace螳 LD1 0.9912. LD1襷 蟆.p <- predict(rs) tmp <- cbind(tmp, p$x) tmp[5:7] plot(rs) plot(rs, dimen=1, type="both")
ろ蠏碁殊朱 蠏碁９覲襦 一危一 覿螳 企讌 .
mean(tmp$LD1[tmp$Species == "setosa"]) mean(tmp$LD1[tmp$Species == "versicolor"]) mean(tmp$LD1[tmp$Species == "virginica"])
蟆郁骸
> mean(tmp$LD1[tmp$Species == "setosa"]) [1] 7.6076 > mean(tmp$LD1[tmp$Species == "versicolor"]) [1] -1.825049 > mean(tmp$LD1[tmp$Species == "virginica"]) [1] -5.78255
鏤